guest wrote:For permanent DVD control, I would have probably built a more complex IR transmitter.
I was impressed with one aspect of the complex design on that site you linked. The maximum safe current in an IR diode depends on duty cycle, so you could design for more than the nominal safe current and depend on the low duty cycle of correct IR signals. But if you did that, then a software error could burn out the diode.
That complex design had a capacitor to avoid RS232 current limiting problems, but that capacitor also acts as a duty cycle limit of the signal. If the software is wrong and the duty cycle too high, that hardware design automatically reduces the duty cycle (the signal won't work but at least the diode won't burn out). So you could tweak the design for higher current and range and still protect the diode.
Unfortunately RSR232 voltages (especially on lap top computers) don't conform to spec very well. That may be a brilliant design for an in spec RS232 port, but for many lap tops it wouldn't work at all (not just the tweaking, the whole design), and for some RS232 ports my idea of tweaking for extra range wouldn't be safe unless components were changed specific to that RS232 port's characteristics.
I'd like to see a design that uses a similar capacitor, but uses the reverse sense of DTR, so it could easily handle a wider range of RS232 ports. That would be software incompatible (the software would need to know the meaning of DTR is reversed) but with so many software options already and open source, what's one more option.
guest wrote:To send the signals (with a current of about 10 mA), I had to put the DVD Player about 2-3 cm in front of the transmitter.
I doubt 10mA would be that bad. I think the RS232 port did not have the characteristics you assumed and the resulting current was much less.
guest wrote:
So I assume, hardware modulation should be prefered. What type of hardware does this job (I'm just interested)?
I forget where I've seen that described. I'd hate to lock in a single frequency as those do. So what design is prefered may depend on how much control you have over what else should running at the moment you want to send an IR command. Maybe the IR transmit software should (and maybe it does) just grab the whole CPU for itself for the time it needs to send the whole command, and just let the sound driver, network driver, and anything else latency sensitive just glitch.